Sanding composites

You can e-mail Steve Beert at As I understand, the sand
blasting prepared the surface for bonding by roughing the epoxy without
cutting and weakening the glass.

Cy Galley - Chair,
Air Emergency Aircraft Repair
A Service Project of Chapter 75
EAA Safety Programs Editor - TC
EAA Sport Pilot
"Peter Dohm" wrote in message
...
I have to presume that he was both quick and gentle with the sandblasting;
as the most common theme that I have heard, over many years, from experts
in
composites is: "the less you sand, the better you sand."

That is very much in keeping with Burt Rutan's remarks, back when he was
actively selling plans for the Vari-Eze and Long-Eze.

The one I remember best was: 'The more time you spend on your cores, the
less time you will spend on your airplane.' The reason for the single
quotation marks is that I believe the statement to be a very close
paraphrase, but probably not exact, as more than twenty years has elapsed.

The other was that a very slightly dry top layer was regarded as less
detrimental than a too wet top layer. Both reduce strength, and the too
wet
surface also adds weight and additional finishing requirements. As I
recall
it, the primary control of the wetness of the layup was the angle of the
squeegee when removing the excess resin which had been stippled in...
Keep
in mind that Burt was quite emphatic regarding the need for the entire
lay-up to be fully wetted out.

Therefore, I really like the idea of a light pass of sandblasting (or bead
blasting) to prepare the surface for the last micro application. I plan
to
use the advice!

"Cy Galley" wrote in message
news:BJxKd.33031$EG1.4274@attbi_s53...
Why sand at all. Steve Beert's prize winning Long was sandblasted, then
micro was squeegee over the top , then a little sanding and paint prep.


wrote in message
oups.com...

Roger wrote:
On 27 Jan 2005 01:04:49 -0800, wrote:

How do you know when you're beginning to sand into the glass cloth?
Do
you simply sand until you
have a smooth finish, then stop? Ot do you keep going?
thanks,

Are you sanding to get a smooth finish or to prepare for the next
lay-up?

If you are getting ready for the next lay-up you are sanding to
"rough
up" the surface to get a good bond to the next layer, not smooth it.

With Vinyl Ester Resin I use 60 or 80 grit (hard to get a smooth
finish with thatG) and just rough up the area. I then vacuum the
area clean and follow up with an Acetone wash. With the area dry or
just barely tacky I do the next lay-up.


For smoothing, as opposed to scuffing, has anyone tried scraping
instead of sanding? Using a scraper like this:

http://www.leevalley.com/wood/page.a...=1,310&p=32669

one can get a finish on wood roughly equivalient to sanding down
to 320 or 400 grit. Scrapers are especially good for removing
bumps and runs in film finishes like shellac and varnish so
I'd imagine that they'd do a good job on hardened epoxies and
resins too.

The cabinet scraper is not a paint scraper, it is a much more
versatile tool. Proper tuning of the scraper is a bit of an
art form but with an understanding of the process and a little
experience one can remove a large amount of material, or only a
small amount with each pass. It is much cleaner and faster than
sanding because it raises shavings instead of making dust.

--

FF

Tim Ward wrote:
wrote in message
oups.com...

Tim Ward wrote:


Well, on sailplanes, anyway, you sand to make sure the airfoil is
the
right
shape, (shrinkage can occur over the spars after some time out of
the
molds), and to make sure that any waves in the contour are less
than
.004
inches.


That's another reason to try scraping. Scrapers can be filed to a
particular curve for just that sort of work.


The curvature changes continuously chordwise, and most sailplane
wings have
taper, so the curvature will change with the spanwise station as
well. So
it's difficult for me to see how you could cut a single curve that
would
match.

You can do that by skewing the scraper and adjusting the angle of
atttack between the scraper and the wing.

How do you vary the curvature when shaping with sandpaper?

Now, for a constant chord wing, that might make an interesting
production
technique:
Build your wing, then build up an extra layer of filler, then
"extrude" the
whole wing panel through a CNC cut scraper, getting exact, smooth
coordinates on the way.


Yes, that would be easier though I still don't see how it would be
done with sandpaper, and think it would be especially difficult
with sandblasting!

It is also the case that not all homebuilt aircraft wings are
fabricated
to the same exacting tolerances as state-of-the art sailplanes, right?
--

FF

On Sun, 30 Jan 2005 21:14:08 GMT, "Cy Galley"
wrote:

You can e-mail Steve Beert at As I understand, the sand
blasting prepared the surface for bonding by roughing the epoxy without
cutting and weakening the glass.

Are you sure he wasn't bead blasting or even CO2 blasting. I've done
a lot of sand blasting and it normally accentuates any imperfections
rather than smoothing. I'd not want to use it on anything other than
steel and even then it will peen the surface causing a warping.

On fiberglass the resin is tough and the glass is hard. If there are
any variations between wet and dry they should show up.

Even the pre molded shells are often sanded through the gel coat and
into the resin. Then given a thin coat of microsphere mix which is
then block sanded with very fine grit.

There have been several G-IIIs at Oshkosh that had the entire wing and
tail assemblies done this way to get an unbelievably true surface. If
I recall correctly the one guy had close to 4,000 hours in prepping
and painting the surface. More than some put into building the same
airplane. that was an all white airplane with a little trim, not like
the lancair 320 that had the Winged goddess on the bottom. I think he
had over 4000 just in the paint and prep.

But for me, and fiberglass... I'd not let any one with a sandblaster
near it no matter how good they are supposed to be. To me, it's just
too risky.

Roger Halstead (K8RI & ARRL life member)
(N833R, S# CD-2 Worlds oldest Debonair)
www.rogerhalstead.com

snip many layers